Mangroves can provide coastal protection by attenuating waves, currents, and trapping sediment (Menéndez et al., 2020; Bao, 2011). The effect of mangroves on the hydrodynamics depends on their size, location, density, distribution and morphology of the vegetation (Mendez and Losada, 2004). Mangrove loss during storm events will therefore impact the coastal protection. The soil-root interaction and influences of soil properties in the resistance of a mangrove tree against failure is neglected in most studies of flood protection by mangroves, due to a lack in understading. This thesis improves the understanding of the soil-root interaction by deriving the schematization for multiple failure mechanisms in comparison with field observations. Firstly, the wind and wave forces are modelled to determine the different loads in the forest. Secondly, a novel schematization was developed of different failure mechanisms considering the soil properties and root system. All results are developed for a case study of a fringe forest in Demak, existing of Avicennia marina rooted in a silty, saturated soil. To be able to determine the difference between the seaward and landward edge of the forest, the width of the forest is increased to 500 meter. Also, due to a lack of information on the mechanical strength of Avicennia m., the mechanical properties of Rhizophora m. are used, which will overestimate the resistance due to the higher wood density of Rhizophora m. (Manguriu et al., 2013). The results show that the drag forces were largely influenced by the tree architecture (such as vegetation width and height), forest density and inundation of the tree. Larger wind and wave forces, and therefore a larger moment occured if the width of the vegetation increased. If the height of the tree increases, a smaller part of the canopy is submerged. This leads to a decreased area that is exposed to waves and therefore a declined wave force. A higher tree does enlarge the area subjected to wind forces, resulting in an increased total moment. Overall, the 2.8-meter tree analysed in this thesis experienced the largest horizontal forces and moment at the seaward edge of the forest due to the relatively large water depth at this location. Furthermore, the maximum horizontal force and moment shift to the landward edge for larger trees due to the increase of wind contribution. To investigate the stability under these loads, the different failure mechanisms need to be inspected. The failure mechanisms of root breakage and slippage are not important because of the high safety factors and no observations in the field. The failure mechanism trunk breakage also showed a high safety factor, contradictiory to field observations pulling willow trees. The model results and field observations showed that a combination of upwind soil uplift and bending of roots was found most likely to occur. For this failure mechanism, the roots are schematized as beams with a spring-support on the leeward side of the trunk. The roots are exposed to the weight of the overlying soil and the overturning moment. The model showed that an increase in the participation angle α or root diameter, or increase in root length results in larger stresses in the windward roots. The amount of uplift enlarges when the root diameter or root length increases or the participation angle decreases. The modulus of subgrade reaction and modulus of elasticity influence the distribution of the overturning moment over the leeward and windward roots. The largest difference between the moment inside the leeward and windward roots is found with a high modulus of elasticity or high modulus of subgrade reaction. This difference is in agreement with the contrasting reaction between soil types as stated in literature (Dupuy et al., 2007). The results show the dependence of the different failure mechanisms on soil and root properties, but also on the forest architecture. This research shows that the incorporation of mangrove stability in wave attenuation models cannot be neglected. Using an effective stress approach, instead of a total stress approach as in this thesis, would provide extra information, such as soil behaviour in time during loading. It is therefore recommended to measure pore pressures under static or cyclic loading, resulting in the progression of the effective stress over time. The inclusion of more accurate soil descriptions would enable reducing uncertainty in nature-based solutions and enables to describe the soil-root reaction to loading over time.

After overtopping, the most common failure mechanism in the world that leads to a dike breach is the instability of the inner slope. Two stability calculations are required for the semi-probabilistic assessment of inner slope stability. One calculation is without significant overtopping and the other calculation is with significant overtopping whereby the dike is saturated. The semi-probabilistic assessment method with overtopping is prescribed in a KPR factsheet, but is not verified by calibration calculations or probabilistic calculations. After performing semi-probabilistic and probabilistic calculations at Jaarsveld-Vreeswijk (JAV) in the Netherlands, there is an inconsistency between the semi-probabilistic and probabilistic estimates of the reliability index for inner slope stability with overtopping. Semi-probabilistic and probabilistic calculations have been performed to gain more insight in the inconsistency between the semi-probabilistic and probabilistic estimates of the reliability index for inner slope stability with overtopping. Based on the results, it is recommended to perform a calibration study for inner slope stability with overtopping to obtain a better estimate of the failure probability based on semi-probabilistic calculations.